Water Resources Management Plan 2019 Annex 8: WRMP strategy · 5 Water Resources Management Plan...

Water Resources Management Plan 2019 Annex 8: WRMP strategy December 2019 Version 1

ii Water Resources Management Plan 2019

Annex 8: WRMP Strategy

Contents

1. Executive summary ............................................................................ 3

2. Investment modelling .......................................................................... 8

2.1 What is investment modelling and why is it used? ........................ 8

2.2 Selection of investment modelling approach ................................ 9

2.3 The Real Options investment model ............................................10

2.3.1 States of the world ................................................................12

2.3.2 ‘Branches’ to represent the range of plausible ‘futures’ .........14

2.4 Modelling inputs and considerations ............................................18

2.4.1 Discount rates and net present value (NPV) .........................18

2.4.2 Option costs and earliest start years .....................................18

2.4.3 Utilisation...............................................................................18

3. Our strategy .......................................................................................20

3.1 Refinement of the strategy from draft to final plan .......................22

3.2 Inclusion of customer preferences ...............................................23

3.3 Environmental cumulative impact assessment and programme appraisal ................................................................................................26

3.3.1 Environmental and social valuation .......................................26

3.4 Regional planning ........................................................................27

3.5 Development of strategies for each area .....................................29

3.5.1 Options and strategy risks .....................................................29

3.5.2 Resilience ..............................................................................30

3.6 Sensitivity testing of the strategy .................................................31

4. References ........................................................................................33

3 Water Resources Management Plan 2019


1. Executive summary In this Annex, we set out the approach we have followed to develop and select the strategy for the

Water Resources Management Plan (WRMP) for each of the company’s three supply areas. The

Technical Overview presents a summary of the process we have followed to develop the WRMP.

Here we focus on the investment modelling part of that process.

The objective of a WRMP is to ensure that there are always enough supplies available to meet

anticipated demands in all water resource zones (WRZs) under every planning scenario or defined

design condition, even under the conditions of greatest water supply stress.

We have developed an economic least cost model (the ‘investment model’) to help select the

combination of options – the portfolio of options – which maintains the supply-demand balance at

least cost. The investment model is a decision making tool that helps the company identify the

optimum set of options based on cost, but it does not necessarily identify the final strategy we adopt

in the plan, as there may be other factors that need to be considered and addressed, such as

customer preferences for different option types, outcomes from environmental assessments of the

options, and regional planning initiatives.

We then also use the investment model to test the robustness of our final strategy against a range

of assumptions, to help identify key alternative options that we may need to investigate in parallel

with the preferred plan.

Separate investment models were developed for each of the three sub-regional supply areas

(Western, Central and Eastern), which are geographically separate (with each supply area consisting

of between three and seven WRZs). Although the building blocks for the strategy are the individual

WRZs, there are inter-connections (either current or potential) between them, and thus actions in

one WRZ can have an impact on other inter-connected WRZs within that sub-regional area. The

model must take account of the supply-demand balances for each planning scenario, including

transfers and bulk supplies, in all the WRZs in each supply area at the same time in order to develop

a consistent solution for the supply area.

The key inputs needed to allow investment modelling are:

◼ The surplus or deficit in supply compared to demand (known as the supply-demand balance)

across the planning period under each of the design conditions and for each WRZ

◼ A list of feasible options (supply-side, demand-side, and water trading) with associated

information on costs and the volume of water these provide in each of the design conditions

There are various different investment modelling approaches that a company could use to support

its decision making for its WRMP. The problem characterisation section of Annex 1 sets out the

company’s analysis for identifying its preferred decision making tool: for WRMP 2019 the company

has employed a Real Options methodology for all three supply areas.

Figure 1 below presents an overview of the inputs and decision making tools.



Figure 1 High level overview of decision making process and inputs

The Real Options approach used to inform the decision making for this plan solves the supply-

demand deficits simultaneously for seven different ‘states of the world’ across five different

‘futures’ or ‘branches’.

◼ ‘States of the world’: represent a snapshot of different climatic conditions and intra-annual

pressures on water resources and demands, from normal year through to severe and

extreme droughts, looking at periods when water supplies are at their minimum, and at

periods of peak water demand during summer months

◼ Different possible ‘futures’ modelled by different ‘branches’: represent a plausible set of

supply-demand balances for a range of possible future scenarios, for which different solutions

may be needed

The use of different futures in the Real Options approach effectively recognises that the future is

not certain, so tries to identify how solutions may change through time in the face of different

possible future water resource pressures, and also identifies a common set of ‘no regrets’ options in

the short term which should be developed regardless of which future may materialise.

These uncertain futures are a key reason why we have adopted the Real Options approach – so that

key schemes and alternatives which address these uncertainties can be investigated and progressed

in parallel to the preferred plan. Should the magnitude of the future uncertainties be less severe,

then some of the schemes would not need to proceed past feasible investigation and planning /

promotion stages. However, the company has little choice but to conduct these investigations of

alternative and preferred schemes through AMP7 (and AMP8), given the scale of uncertainties the

company faces in the next 10 years.



Our approach to developing the strategy for the draft WRMP in each of our supply areas is

summarised in Figure 2 below. The first stage was to undertake an initial phase of scenario testing

to help understand the sensitivity of the strategy to various possible constraints. The purpose of this

testing was ultimately to inform the selection of the company’s plan. This stage involved an initial

‘least cost’ model run to develop a ‘basic solution’, without further consideration of potential

constraints.

This was then tested by, for example, modifying assumptions about the availability of certain options

to progress our understanding of the impacts that these assumptions might have on the strategy.

From examination of the various model run outputs, and taking into account the pre-consultation

discussions with regulators and stakeholders, consultation representations, and policy decisions,

refinements were introduced to reflect a ‘constrained’ least cost strategy. The policy decisions

applied were in regard to the inclusion of water efficiency assumptions (the company’s target to help

our customers achieve an average per capita water consumption of 100 litres per person per day by

2040 – the ‘Target 100’ policy), the policy of leakage reduction (aiming to achieve a 15% reduction

by 2025 and 50% reduction by 2050) and the availability of Drought Permits / Orders to relax

abstraction licence conditions in severe and extreme drought events.

The constrained least cost strategy was then examined and tested against:

◼ Environmental criteria

◼ Outcomes from regional planning exercises (Water Resources in the South East – WRSE

◼ The preferences arising from customer engagement activity

The testing of the constrained least cost strategy against the environmental, regional and customer

preferences criteria effectively led to an iterative loop. Another key element considered was the

relative impact of the changes influenced by testing against criteria in terms of the overall strategy

cost, compared to the least cost model and to the constrained least cost strategy. Where there is

little cost difference, and the change of option provides a more positive outcome to one or more of

the testing criteria, then there is a stronger case for including the option change in the preferred

strategy.

Following this review and testing process, any refined assumptions of the feasible options set were

fed into the Real Options model to derive the strategy. The strategy was then subjected to scenario

and sensitivity testing to understand what alternative strategic schemes may be needed, should

it not be possible to implement the schemes in the preferred plan. This is particularly important for

those schemes in the strategy that are required early in the planning period, in AMP7 or AMP8.



Figure 2 Development of strategy for the draft WRMP

The draft WRMP strategy is published for consultation with customers, stakeholders and regulators.

The responses received during consultation may result in changes to the assumptions or inputs used

to derive the supply-demand balances, as well as to the set of options that are available to meet

forecast deficits. Consequently, elements of the strategy may be reviewed or refined as we move

from the draft plan to our final plan. This process is summarised below in Figure 3.



Figure 3 Development of the strategy from draft to final plan



2. Investment modelling

2.1 What is investment modelling and why is it used? The objective of a Water Resources Management Plan (WRMP) is to ensure that there are always

enough supplies available to meet anticipated demands in all water resource zones (WRZs) under

every planning scenario or design condition, even under the conditions of greatest water supply

stress, throughout the planning period. The planning period we are focused on for the WRMP is from

2020 to 2070.

Therefore, where we have identified that there is a supply-demand balance deficit for a WRZ during

the planning period, we must identify what options can be developed to address that deficit. To do

this, we use an economic least cost model (the “investment model”) to help select the combination

of options – the portfolio of options – which maintains the supply-demand balance at least cost

(discounted), given the assumptions for each planning scenario.

The investment model is a decision making tool that helps the company identify the optimum set

of options based on cost, but it does not necessarily identify the final strategy we adopt in the plan,

as there may be other criteria that need to be considered and addressed to identify the final strategy.

These criteria may include customer preferences of different option types, outcomes from

environmental assessments of the options, and regional planning initiatives. We then use the

investment model to also test the robustness of our final strategy, and to identify key alternative

options that we may need to investigate in parallel to the preferred set of options.

Separate investment models were developed for each of the three sub-regional supply areas

(Western, Central and Eastern), which are geographically separate. Each supply area consists of

between three and seven WRZs. Although the building blocks for the strategy are the individual

WRZs, there are inter-connections (either current or potential) between them, and thus actions in

one WRZ can have an impact on other inter-connected WRZs within that sub-regional area. The

model must take account of the supply-demand balances for each planning scenario, including

transfers and bulk supplies, in all the WRZs in each supply area at the same time in order to develop

a consistent solution for the supply area.

The investment model incorporates all the feasible supply-side, water trading and demand-side

options. These were all made available within the model to solve the supply-demand balance deficit

in each of the planning scenarios.

Existing inter-company bulk imports and exports were included within the investment model as ‘fixed’

baseline transfers, so they were fixed assumptions in the supply-demand balance that the model

aimed to solve. The reason for this is that bulk supplies are treated as contractual volumes of water

to be supplied, except where the severity of the drought affects the ability of the donor company to

provide the contractual amount, and some form of ‘pain share’ is introduced. It is therefore prudent

to plan on the basis of meeting the bulk supply commitments. This is explicitly stated in the

company’s Drought Plan. New water trading options are treated like any other option – the model

may select them freely to solve the planning problem. The volumes associated with the new water

trading options were also ‘fixed’ as a contractual volume of water, and these volumes were agreed

in discussion with the donors/receivers under each of the relevant planning scenarios.

Inter-zonal transfers were treated differently within the model. These are internal transfers between

Southern Water’s own WRZs. As such, all inter-zonal transfers were included as options for

transferring between zones and were then selectable as part of the optimisation process. This is

because an internal transfer does not affect the overall water balance for the supply area; they are

just a different way of balancing the water available across the inter-connected WRZs. The model



was therefore allowed to vary transfers from zero up to the capacity of a given transfer within the

optimisation process to derive an optimal least cost solution. This allowed the model to select the

least cost overall strategy for all water trading and transfers, resource development, catchment

management and demand management options.

The key inputs needed to allow investment modelling can thus be summarised as follows:

◼ The surplus or deficit for the supply-demand balance across the planning period under each

of the design conditions and for each WRZ

◼ A list of feasible options with associated information on costs and volume of water these

provide in each of the design conditions

2.2 Selection of investment modelling approach There are various different investment modelling approaches that a company could use to support

its decision making for its WRMP (UKWIR, 2016). As described previously in Annex 1, the company

has set out its approach to deciding on its preferred decision-making approach. Two potential options

were identified from the problem characterisation assessment; these were:

◼ Real Options using modified EBSD (Economics of Balancing Supply and Demand) models

◼ Adaptive pathways

The adaptive pathways methodology is most closely aligned with the company’s adopted ‘risk

principal’, preferred approach and key challenges. There are a number of key uncertainties that are

yet to be resolved either in the magnitude of their impact on our supply-demand balance, the timing

of impacts, or both. An adaptive pathways methodology would explicitly recognise that such

uncertainties exist and track them. As the uncertainties are resolved over time, the plan and

subsequent future actions could then be appropriately adapted to ensure an optimal cost-benefit

plan. However, as noted by UKWIR (2016) this methodology has not yet been applied to a water

resources problem and hence the techniques remain untested and highly uncertain. Therefore, the

company intends to instigate a research project to further investigate how an adaptive planning

methodology can be developed and applied in the context of water resources planning in the United

Kingdom with an aim of further developing the approach in our next WRMP (known as WRMP24).

In AMP5 the company tested a pilot planning approach using a Real Options methodology to

examine the potential schedule of investment options associated with uncertain sustainability

reductions in the Western area in order to establish ‘no regret’ investments, although this work was

not completed until after the WRMP14 planning was undertaken, so was not part of the WRMP14

submission. For this current plan the company decided that it will employ the Real Options

methodology across all three supply areas.

The company also identified a preference to develop a multi-criteria analysis (MCA) approach to

developing a best value plan alongside the Real Options methodology. This considers, in particular,

environmental assessment outcomes, regional planning and customer preferences for different

option types, with the objective to derive a best value plan from the initial least cost solution. We are

looking to further develop and refine this ahead of the next WRMP in 2024.



2.3 The Real Options investment model Our WRMP14 used a conventional investment modelling approach in line with the EBSD

methodology, with sensitivity analysis and stress testing to inform the selection of the company’s

preferred plan.

The investment model used for WRMP14 comprised a Mixed Integer Linear Programme to optimise

the selection of options to meet a specified deficit on a least cost basis. Thus the WRMP14

investment model had to solve a single objective – namely least cost (which included financial,

environmental and carbon costs) – whilst satisfying a single constraint to ensure that supplies would

be greater than demand plus headroom in each year of the planning horizon under different ‘states

of the world’ (which were used to provide better utilisation calculations relevant to variable operating

expenditure, or variable ‘opex’).

There are limitations to using a conventional EBSD (and also to using a more simplistic Average

Incremental Cost, AIC) approach, particularly where there are substantial uncertainties over time

– which is the case Southern Water face, particularly during the 2020s with sustainability reductions,

and beyond that with demand growth and climate change impacts on supplies. In such cases, a Real

Options decision making approach is considered a better tool, as it provides a framework for

improved learning about, and consideration of, the uncertainties, and thus the flexibility to adapt

plans to reflect this learning. Real Options approaches are most appropriate where the level and

nature of uncertainties at different points in the future could change the optimal investment decisions

today1. The real options approach concentrates on understanding what a company has to do in the

near term, against the context of what might happen in the future – i.e. it aims to identify a ‘least

regrets’ solution set.

A real options approach incorporates planning for a wider range of plausible futures compared to the

previous EBSD approach and thus aims to ensure that the plan is more resilient against a range of

uncertain, yet possible, futures, in addition to allowing development of greater plan flexibility.

Southern Water undertook a pilot study at the end of AMP5 to demonstrate the principles of Real

Options and multi-criteria analysis. This was not published as part of WRMP14. The investment

modelling for this plan has therefore built on the work undertaken in AMP5, as well as utilising current

industry best practice in terms of:

◼ UKWIR (2016), WRMP19 methods – decision making process, 16/WR/02/10

◼ UKWIR (2016), WRMP19 methods – risk based planning methods, 16/WR/02/11

Figure 4 sets out a high level overview of the decision making approach followed for this plan. There

are a number of models used to generate the required supply-demand balances that are then used

as inputs to the Real Options model. The outcome of the options appraisal process – a feasible list

of costed options with associated volumes under different drought return periods – provides the other

key input to the Real Options model.

1 UKWIR (2016) WRMP 2019 Methods – Decision Making Process: Guidance



Figure 4 High level overview of decision making approach, inputs and tools for this plan

The Integrated Risk Model and Scenario Generator Model are used to generate the supply-

demand balances that need to be solved by the Real Options model, as described more fully in

Annex 5.

The feasible list of options with costs, earliest start dates, etc. are generated from the options

appraisal process, described in Annex 6.

The Real Options approach used to inform the decision making for this WRMP solves the supply-

demand deficits simultaneously for seven different ‘states of the world’ across five different

‘futures’ or ‘branches’.

◼ ‘States of the world’: represent a snapshot of different climatic conditions and intra-annual

pressures on water resources and demands, from normal year through to severe and

extreme droughts, looking at periods when water supplies are at their minimum, and at

periods of peak water demand during summer months

◼ Different possible ‘futures’ modelled by different ‘branches’: represent a plausible set of future

supply-demand balances for a range of possible future scenarios, for which different solutions

may be required

The investment decisions are optimised to ensure we can meet our target level of service across a

range of drought severities at different times of the year, whilst still considering the operation of

schemes during normal climatic conditions. The use of different futures in the Real Options approach

effectively recognises that the future is not certain, and so the method tries to identify how solutions

may change through time in the face of different possible future water resource pressures. The



approach therefore tries to ensure that the plan is resilient against a range of uncertain, yet possible,

futures that the company may face. This is particularly important where the scale of the uncertainties

is large (for example from potential ‘sustainability reductions’ of licensed abstractions). The objective

of our approach is therefore, to ensure that the plans cover a wide, yet appropriate, range of futures

to ensure that all the key strategic options are identified. This is critical because there may not

otherwise be sufficient time from when the sustainability reductions are confirmed for implementation

to develop appropriate schemes. These uncertain futures are a key reason why we have adopted

the Real Options approach – so that key schemes and alternatives which address these

uncertainties can be investigated and progressed in parallel to the preferred plan. Should the

magnitude of the future uncertainties be less severe, then some of the schemes would not need to

proceed past feasible investigation and planning / promotion stages. However, the company has

little choice but to conduct these investigations of alternative and preferred schemes through AMP7

(and AMP8), given the scale of uncertainties the company faces in the next 10 years.

The plan is tested against a range of criteria to ensure that it represents best value, not just least

cost. This included inputs from regional planning exercises, customer preferences for different option

types, and environmental considerations informed by the Strategic Environmental Assessment,

Habitats Regulation Assessment screening, and Water Framework Directive assessments. This is

discussed further in the section 3 below.

We have also included a ‘conventional’ EBSD assessment alongside the Real Options approach to

provide a comparison between the approach adopted previously for WRMP14 (the EBSD approach),

and the new approach for this plan. This is discussed further in the section 3 below.

2.3.1 States of the world

The supply-demand balance will vary throughout the year, as available supplies and customer

demands for water fluctuate. This ‘within year’ variability highlights the need for assessment of a

number of different design conditions or ‘planning scenarios’ that must be considered for each year

of the planning horizon. As described in Annex 3, the planning scenarios that the company looks at

are:

◼ The annual average (AA) – which may also be referred to as the average deployable output

(ADO) planning scenario. This scenario compares the average daily demand over the year

against the average daily supplies that are available over the year

◼ The critical period (CP) – corresponds to the period of peak water demand, which

normally occurs during the summer months of June, July and August. The peak period of

demand is generally defined in terms of the average day peak week (ADPW) demand. The

peak demand is compared to the supplies available during that same summer period. This

may also be known as the peak-period deployable output (PDO) planning scenario. During

these summer peak periods, it is generally not the availability of water resources to meet

peak demands that is the constraining factor, but the capacity of the infrastructure

◼ The minimum deployable output (MDO) period – used to assess the period where available

supplies are expected to be at their lowest or most stressed – i.e. it represents the ‘minimum

resource period’. This MDO period normally occurs during late summer/early autumn when

river flows are at their minimum following the summer, and groundwater levels are at their

lowest prior to the onset of winter recharge. The demands under this scenario are based on

the minimum rolling 30-day average daily demand over the same relevant period

The Western area is most susceptible to the ‘minimum resource period’ and to the critical period (i.e.

peak summer demand period), as the two largest WRZs within the area have a significant proportion

of supplies from run-of-river abstractions and there is no surface water reservoir storage.



The Central area is similar to the Western area, as the Sussex North WRZ takes a significant

proportion of its supply from a run-of-river source although there is a surface water storage reservoir.

The two Sussex coastal WRZs are supplied entirely by groundwater, although there is also a transfer

from Sussex North to Sussex Worthing.

The Eastern area differs from the other two supply areas, as the Bewl-Darwell reservoir system

provides the ability to manage seasonal drought events. Hence the Eastern area is most susceptible

to the annual average and critical period (summer demand) planning scenarios.

The various states of the world allow differing drought conditions to be considered in combination

with inter-annual variability in supplies available to meet demand for water. Each state of the world

therefore has its own supply-demand balance – i.e. its own profile of surpluses or deficits over the

planning period. The model must solve each of the states of the world simultaneously (ie so

that any deficit in any state of the world is solved).

Inclusion of the states of the world is useful for a number of reasons:

◼ It ensures that the plan is robust against a range of supply and demand conditions that could

be faced by the company in any given year across the planning horizon

◼ It allows consideration of how the water available from different options may vary in different

drought events

◼ It allows additional drought intervention options to be considered alongside the water

resources options in more extreme droughts

◼ It ensures that the costs are appropriately weighted in relation to how options are likely to be

used under each state of the world (known as ‘utilisation’ – see section 2.4.3). Hence an

option that is only required to meet an extreme event is likely, on average, not to have

significant total variable operating costs, as it would only be required to supply water very

infrequently (note that the capital costs of the option and any fixed operational costs would

still need to be paid for regardless of how frequently the scheme may actually be used in

practice – i.e. the capex (capital expenditure) and fixed opex are independent of the

utilisation)

The states of the world are related to the following climatic conditions, or design drought events:

◼ Normal year – 50% annual probability – relating to typical non-drought climatic conditions,

with average customer demand

◼ Drought condition – a 1 in 20 year drought, or 5% annual probability

◼ Severe drought condition – a 1 in 200 year drought, or 0.5% annual probability

◼ Extreme drought condition – a 1 in 500 year drought, or 0.2% annual probability

For each of these drought climatic conditions there is a state of the world for each of either the

minimum resource period or annual average (depending on the supply area we are modelling)

and for the peak demand period planning scenarios. The exception to this is for the normal year,

for which there is not generally a deficit. Under this condition only the annual average period is used

(not the critical period). The inclusion of the normal year annual average state of the world ensures

the appropriate calculation of variable costs based on expected utilisation (as explained in section

2.4.3).

Section 39B(2) of the Water Industry Act, requires the company when planning for drought, to plan

to supply adequate quantities of wholesome water, with as little recourse as reasonably possible to

drought orders or drought permits. In ensuring compliance with this, previous Water Resource

Planning Guidance (WRPG) only required planning to be based on the worst historic event and water

resource planning was not required to take into account wider severe drought conditions. The WRPG



for this plan (Environment Agency and Natural Resources Wales, 2017) has changed to now

recognise the need for resilience in a severe drought condition (a 1 in 200 year drought event). Our

previous WRMP14 already planned to a severe drought (1 in 200 year drought event) without any

recourse to drought permits and orders. Planning in line with the WRPG therefore already

reflects a continuation of our level of service. We have therefore chosen our States of the World

to carefully reflect the levels of service.

However, in this plan, we have also sought to understand the impacts of more extreme drought

events (1 in 500 year drought event), as this aligns with the latest thinking around drought resilience

(e.g. as reported in the recent National Infrastructure Commission report which highlighted the need

for increased drought resilience to reduce or minimise the significant economic impacts of ‘level 4’

drought restrictions (stand pipes and rota cuts)).

In line with our continued practice of moving water resource planning forward, we have only allowed

drought permits and orders to be selected in the investment model in an extreme drought event (1

in 500 year drought event) so as to ensure that the WRMP can be resilient to a level in line with

guidance, in line with our levels of service and in line with the requirement to plan with as little

recourse as reasonably possible to drought orders and drought permits. It also means that the

selection does not drive excessive infrastructure; but it still allows a progressive and pragmatic

approach to exploring extreme drought events.

It is important to recall that all the states of the world must be solved simultaneously in the Real

Options model. What we are examining when we look at both the severe and extreme states of the

world is thus the balance in the solutions between the portfolio of options needed in severe droughts

without drought interventions (except in the short term), with that same portfolio of options in

combination with drought interventions in extreme droughts. We are effectively examining whether

we have sufficient options to meet differing levels of drought when considering that drought

interventions would be used in more extreme droughts. But we are also recognising that these

drought interventions may not be available in all WRZs in a supply area, and that the existing

connectivity between WRZs may be limited. Our analysis therefore considers the resilience of

transfers between the WRZs, and the potential need for increased connectivity.

Our sensitivity testing includes a scenario where the drought permits and orders are not available in

the extreme drought to demonstrate the impact this assumption would have on the options to

maintain the supply-demand balance. We have also included a sensitivity test where we remove the

extreme drought states of the world from the Real Options model, to demonstrate the impact that

solving for extreme droughts has on the plan. This effectively helps to identify whether the severe or

extreme droughts are driving the investment in water resources options.

2.3.2 ‘Branches’ to represent the range of plausible ‘futures’

The Real Options approach is used to understand how a plan would vary against a range of future

scenarios that result from uncertainty about forecasting the future. Despite the uncertainties with

different possible futures, the WRMP must present a preferred set of options, and as a result a

number of schemes may be needed in the short term before the uncertainties have been better

understood. The company wants to ensure that the plan for the short term is flexible enough to be

optimal against a wide range of possible, yet plausible, futures.

The Real Options method seeks to provide the flexibility to act on new information that becomes

available over time, whilst ensuring that schemes that are needed in the near term are implemented

and do not become rapidly redundant (i.e. it seeks to develop a ‘no regrets’ solution). This is the key

component of a Real Options approach; it effectively recognises that the future is not certain, and so



it tries to identify how solutions may change through time in the face of different possible future water

resource pressures.

The different futures (also referred to through this Annex as ‘branches’) are built up from a

combination of possible scenarios relating to demand growth, climate change impacts on water

supplies, and sustainability reductions (changes to the licensed volumes of water that a water

company is authorised to abstract, with the aim of ensuring that the abstraction does not pose an

unacceptable risk to the water environment).

The baseline supply-demand balance forecast is generated as a series of probability distributions

from which the company can select different percentiles to represent a range of possible futures. As

the branches are a combination of the probability functions of the three key uncertainties it is not

possible to say what any given branch represents (i.e. you cannot say that branch A is a high

sustainability reduction scenario, branch B high demand growth, etc.) These supply-demand

balances are used as the input to the Real Options decision making model with selected percentiles

making the ‘branches’ of the Real Options model.

The development of the branches and their underlying assumptions and generation of the

subsequent range of supply-demand balances (surpluses or deficits over the planning period) for

each of the futures is described in Annex 5.

An indicative Real Options branch plot (described more fully in Annex 5) is shown in the figure below,

which demonstrates the range of possible supply-demand futures of the five branches.

The supply-demand balances used as the ‘futures’ or ‘branches’ in the Real Options model reflect

the following percentiles2:

◼ 10th percentile (larger deficits)

◼ 30th percentile

◼ 50th percentile (the middle branch – representing the more traditional supply-demand balance

that would have been investigated through a traditional investment modelling approach)

◼ 70th percentile

◼ 90th percentile (smaller deficits, or in surplus)

2 For the draft WRMP, the Western area was treated differently to the Central and Eastern areas, because the scale of sustainability reductions faced on the rivers Test and Itchen was considered to far outweigh the other uncertainties. However, following the conclusion of the River Itchen, River Test and Candover abstraction licence Public Inquiry in March 2018, the scale and timing of these two sustainability reductions was clearer. We have subsequently had our licences changed (in March 2019). Therefore, the approach in the Western area has been revised to align with the Central and Eastern areas, so including the remaining uncertain sustainability reductions with demand growth and climate change impacts on supplies probabilistically to develop a range of plausible futures.



Figure 5 Indicative branching of Real Options for different percentiles of supply-demand balance

We assume that there is a single branch in the short term, based on the middle branch – the 50th

percentile. This allows a common solution in the short term. The branches then diverge at a

branching point, which for our WRMP, was selected to be 2027. The selection of this branching point

was based upon a number of factors, primarily it was the point at which the large uncertainties

relating to potential sustainability reductions are likely to be realised (as discussed in Annex 3). The

sustainability reductions could represent a significant step change in the supply-demand balance,

and hence it seemed an appropriate point to explore how the solution would vary from this point

onwards in different futures.

However, any solution that is needed at, or just after, the branching point, will actually have a

decision point earlier in planning period, due to the need for planning and promotion activity and

construction of the schemes. Hence, some schemes may not actually be needed in lower deficit

branches, but because of the time required to plan, develop and build them, they will show up in all

5 branches at the decision point.

The statutory process for WRMPs requires them to be reviewed and updated at a minimum every 5

years. At the point when the next WRMP is prepared, it is expected that some of the uncertainties

around the impact of sustainability reductions (in particular) will be better understood. By this point

some of the schemes identified for implementation in 2027 may not be required or may be deferred

until later in the planning period. In such cases, the scheme would have already gone through a

certain amount of promotion, investigation, and planning. But no further activity would be taken on

these tasks for the scheme once it was identified that it was no longer required.

However, if at the point of reassessment in AMP7, the future does look more like one of the higher

deficit branches we have identified in our current plan that have driven the selection of some of the

schemes in 2027, then the work in early AMP7 on promotion and planning will have enabled the

scheme to be successfully implemented in 2027, to address the supply deficit. If the promotion and

planning activity had not taken place in AMP7, there would be a significant risk that the scheme

could not be implemented when needed in AMP8.

This is summarised in the indicative diagram in Figure 6 below.



Figure 6 Process of scheme selection and development under different branches

This is a critical aspect of the real options approach. In summary, with reference to the example in

Figure 5:

◼ By the middle of AMP7, as part of the company’s preparations for its next plan (WRMP24),

some of the uncertainties we currently have about the future (in the short to medium term)

may be better understood. For example, environmental investigations may be completed and

so the scale of sustainability reduction impacts better understood. Hence the company should

have a clearer indication as to which of the possible supply-demand balance future branches

is most likely

◼ In the meantime, the company will have been able to start to implement the ‘no regrets’

options required in the short term, and undertake feasibility investigations and planning

preparations for the next set of options required in AMP8 under the different potential futures

As the potential ‘futures’ are selected from the probabilistic combination of the scenarios, it is

not possible to identify exactly what is contributing to a given future, as represented by one of

the five percentile branches. The key point is that the branches represent plausible potential

future deficits in the face of uncertainty, and we try to solve these, without needing to know

exactly what might be driving the future deficit. We have purposefully not chosen the most

extreme combination of futures (which would represent the worst case for all of the drivers

combined); instead we have curtailed the selection to ‘plausible’ futures within the 10th and 90th

percentile ranges.

For our main model runs, we have assumed that each future is equally plausible, as we do not have

any evidence to the contrary. The cost of solving each branch is therefore weighted equally. We

have, however, included different assumptions for sensitivity testing of the preferred plan, where

different weightings have been applied to each branch, to understand the impact that this can have

on the plan. These sensitivity tests are presented in Annexes 9-11.



2.4 Modelling inputs and considerations 2.4.1 Discount rates and net present value (NPV)

Discounting is a technique that is used to compare costs and benefits that occur in different time

periods. It is based on the principle that people generally prefer to receive goods and services now

rather than later, which is a concept known as ‘time preference’. For individuals, time preference can

be measured by the real interest rate on money lent or borrowed. The concept can be expanded to

society as a whole; where there is also a general preference to receive goods and services sooner

rather than later, and to defer costs to future generations. This is known as ‘social time preference’.

The ‘social time preference rate’ (STPR) is thus the rate at which society values the present

compared to the future (HM Government, 2003).

When undertaking an economic appraisal of feasible options over the entire planning period it is

important to account for this social time preference. Discounting is therefore the process by which

all the future costs and benefits are converted to today’s (present) value, so that they can be

compared on a like-for-like basis. By summing the present values of all the costs and benefits

associated with an option, the Net Present Value (NPV) (of both costs and benefits) of that option is

derived. This is used to compare all feasible options in the investment model, and hence to derive

the least cost set of options which are needed to meet a given supply-demand balance deficit.

All costs and benefits in the options appraisal and least cost economic model were discounted

according to the time-varying Treasury Green Book discount rates, i.e. 3.5% for years 0-30 of the

appraisal period, 3% for years 31-75, and 2.5% for years 76-125, as specified in the EA’s WRPG.

2.4.2 Option costs and earliest start years

Annex 6 describes the information compiled for feasible options including the development of costs and earliest start years for implementation of the schemes. Options which could have a variety of sizes (e.g. a desalination plant could be 10Ml/d, 20Ml/d or

30Ml/d) are represented in the Real Options model as modular components. This provides flexibility

for the model to select increasing increments of such schemes – i.e. 0 to 10Ml/d in year x, then

increasing from 10 to 20Ml/d in year y, etc.

2.4.3 Utilisation

All schemes and interventions in the model can be categorised into those that could be turned off

when they are not required or run at a minimum capacity (e.g. desalination, water reuse), and those

that are maintained at a constant output (e.g. leakage). For the former set of schemes, a ‘utilisation

factor’ is applied to each of the states of the world to reflect the reduced operational costs that could

be obtained by only operating the schemes when they are needed (i.e. during the lead in to and

during design drought conditions).

The utilisation factors are only applied to the variable operating costs (opex) and not to any fixed

costs. This is because fixed costs (capital costs and fixed opex) must be incurred if the option is

implemented, regardless of how much that option is actually used. However, for variable opex, which

is generally expressed as a cost per unit of water produced, the cost has to be weighted to reflect

the proportion of the year covered by each ‘seasonal’ condition (critical period or minimum resource

period), as well as the different drought design conditions that are examined. The states of the world

therefore have different utilisation factors to reflect this cost weighting.

The Real Options investment model includes utilisation within its calculations by solving

simultaneous supply-demand balances across all the WRZs in each area for all the states of the

world. The utilisation factor that is assigned represents the frequency with which a given state of the

world is expected to occur.



The investment model is set up to minimise the fixed and variable opex costs while satisfying all the

supply-demand balances and their respective utilisation factors. The optimiser thus seeks the overall

least cost for the plan that allows for the expected amount of time that the system would operate

under each state of the world throughout the planning period. Under most circumstances new

schemes do not have to be operated during ‘normal’ climatic conditions, so the scheme only incurs

the fraction of the variable opex costs that would be required to meet the drought conditions under

the various planning scenarios. (Note, however, that some schemes are assumed to be run

continuously at a lower rate regardless of whether they are needed to solve a deficit in a given state

of the world; so will still incur some variable opex in states of the world when they are not actually

required to solve a deficit – e.g. desalination options have been assumed to be run at a minimum of

25% capacity in all time periods).

The design drought conditions each have an assigned return period, which is converted into a

probability – i.e. the severe drought is up to a 1 in 200 year event, which equates to 0.5% probability.

The ‘normal year’ condition was equal to the 50% probability. The drought utilisation is calculated as

the change from moving from one drought state to the next – for example, the extreme drought

utilisation recognises that any event exceeding the 1 in 200 year event will effectively be an extreme

drought condition, so the drought utilisation was calculated as an average of the two drought

probabilities.

The utilisation factors used for each state of the world must then take account of the ‘within year’

variability, which reflects that the supply-demand balance will vary throughout the year as available

supplies and customer demands for water fluctuate. We have assumed that the annual average or

minimum resource period will be the representative level on which to base utilisation for three-

quarters of the year, while the critical period peak demand period will cover the remaining three

month period (i.e. the summer months June to August). However, the peak demand is not expected

to last for the whole three-month period, so a peaking factor is used to account for this.

Table 1 below shows the derivation of the utilisation factors as applied to each state of the world.

Table 1 Derivation of utilisation factors for the states of the world

Name Return

Period Probability

Drought

utilisation

Utilisation

(CP)

Utilisation

(MDO/AA)

Normal year 2 0.5 0.694 n/a 69.40%

Drought 20 0.05 0.275 6.21% 20.72%

Severe drought 200 0.005 0.0275 0.62% 2.07%

Extreme drought 500 0.002 0.0035 0.08% 0.26%

The Real Options model has also been developed to be able to assign a probability to each of the

potential futures or branches to represent the perceived likelihood of that future. This probability is

applied as an expected cost weighting to the total cost calculation. We have assumed that each

branch will have an equal probability, because there was little information on which to base an

alternative weighting scheme.



3. Our strategy The plan has been formulated through an iterative process of economic least cost modelling. The

ultimate objective of this process is to define a strategy, comprising a portfolio of schemes that:

◼ Provides secure supplies of water

◼ Protects the environment

◼ Represents best value for customers and reflects their preferences

Our approach to developing a strategy for the draft WRMP in each of our supply areas is summarised

in the figure below.

Figure 7 Development of strategy for the WRMP

The first stage was to undertake an initial phase of scenario testing to help understand the sensitivity

of the strategy to various possible constraints. The purpose of this testing was ultimately to inform

the selection of the company’s plan. This stage involved an initial ‘least cost’ model run to develop

a ‘basic solution’, without further consideration of potential constraints.

This was then tested by, for example, modifying assumptions about the availability of certain options

to progress our understanding of the impacts that these assumptions might have on the strategy.

From examination of the various model run outputs, and taking into account the company’s policies,

business planning decisions, and pre- and post-consultation discussions with regulators and

stakeholders, policy decisions and refinements were introduced to reflect a ‘constrained’ least cost

strategy. The key policy decisions applied were in regard to the inclusion of water efficiency

assumptions (the company’s target to help our customers achieve an average per capita water

consumption of 100 litres per day by 2040 – the ‘target 100’ policy), the policy of leakage reduction

(aiming to achieve a 15% reduction by 2025 and 50% reduction by 2050) and the availability of

Drought Permits / Orders to relax abstraction licence conditions in severe and extreme drought

events.



The constrained least cost strategy was then examined and tested against a number of criteria:

◼ Outputs from environmental assessments of the options. To address whether the

combination of options and timing of the need for them present significant risks or have

planning and promotional issues that might affect the deliverability of a scheme or schemes.

This represents a second stage of the environmental screening process included as part of

the options appraisal process, to develop a feasible set of options; however, timing of option

implementation and cumulative impacts are clearly important additional considerations, as

well as feedback from consultation responses on certain options

◼ Outcomes from regional planning exercises (Water Resources in the South East – WRSE).

A cross-check against the outputs from the WRSE modelling scenarios and review against

bi-lateral discussions held with neighbouring water companies covering bulk supply needs

and timing, schemes that could be jointly developed, the reliability of the bulk supply to the

different drought design conditions, and the costs associated with the development of

sources that support a potential new bulk supply

◼ The preferences for different option types arising from customer engagement activity.

To reflect the ranking of option types from customer preferences surveys conducted during

pre-consultation and during the consultation period. This also takes account of consultation

responses on specific options.

Note that overlaying the Strategic Environmental Assessments (SEA), regional planning and

customer preference considerations on the constrained least cost strategy does not necessarily

mean it will need to be changed – i.e. it may already adequately address key considerations from

these criteria. Additionally, although some schemes may be less favoured by the SEA, regional plans

or customers, the availability of suitable, better alternatives, or the deficit faced may mean that some

options need to be retained in the feasible list regardless. It is also possible that these criteria could

sometimes contradict each other – e.g. a scheme identified from WRSE may not align with, say,

customer preferences; in which case, the company must exercise its judgement to weigh the pros

and cons of a given scheme and the alternatives that would otherwise be needed. This represents

a process of qualitative multi-criteria assessment.

The original intention for applying multi-criteria analysis had been to develop monetised costs or

‘penalty values’ to allow the Real Options model to take account of these preference costs in a

quantitative way along with the feasible options costs, when optimising the plan against a least cost

objective.

However, at the time of formulating the draft WRMP, there was limited information available on which

to develop suitable shadow costs (i.e. penalty / reward functions linked to environmental metrics) to

apply to the outputs from the environmental assessments SEA, Habitats Review of Abstractions

(HRA) and Water Framework Directive (WFD). We will continue to develop the Natural Capital

approach to determine if this technique can help provide a better way of determining the penalty cost

function for the environment. As described below in the section 3.3.1, there are also potential issues

and limitations with monetising the environmental and social costs/benefits for each option directly.

It was therefore decided that the best approach for this plan was to use the outputs from the

environmental assessments (without any attempt to monetise them) applied to the initial

‘constrained’ least cost strategy, to inform the conclusions on the preferred portfolio of options in the

strategy.

There was a similar rationale with regard to the inclusion of customer preferences values. A number

of different ways were selected to derive the preference costs based on YouGov surveys (see Annex

1). Whilst the customer preference surveys did elicit some monetary values, the unit costs for

different option types were not available to use for this plan – further work is being undertaken on

this. In order to include allowances of customer preferences, we reviewed the customer preferences



from the surveys and engagement activities, and reviewed the options selected in the constrained

least cost model to identify preferential option types, where there were suitable alternatives available.

This helped the company to decide on the option types that should be included in the strategy, and

also where other option types should be excluded or their use minimised. However, it may not be

possible to exclude a least favoured option entirely, if there are limited alternative options to solve a

given supply-demand balance deficit. We intend to undertake further work to develop suitable

penalty functions based on customers’ preferences for different types of options for the next AMP

(for WRMP24).

The process of testing the constrained least cost plan against the environmental, regional and

customer preferences criteria was therefore iterative. The other key element considered was the

relative impact of the changes influenced by testing against criteria in terms of the overall strategy

cost, compared to the least cost model and to the constrained least cost strategy. For example,

where there is little cost difference and the change of option provides a more positive outcome to

one or more of the testing criteria, then there is a stronger case for including the option change as

part of the strategy.

Following this review and testing process, any refined assumptions of the feasible options set were

fed into the Real Options model to derive the strategy for the WRMP.

The strategy for the WRMP was then subjected to scenario and sensitivity testing to understand

what alternative strategic schemes may be needed, should it not be possible to implement the

schemes in the preferred plan. This is particularly important for those schemes in the strategy that

are required in AMP7 or AMP8.

Where there is uncertainty around the delivery of these schemes, the company may need to conduct

feasibility investigations of alternative schemes (and potentially environmental surveys and planning

activities) in parallel to developing the portfolio of schemes selected in the strategy, to ensure that

there will be a scheme in place to solve an identified future deficit.

For comparative purposes, we have also undertaken investment model runs using a conventional

EBSD approach (which was the approach used in the previous WRMP published in 2014). The area-

specific detail provided in Annexes 9 to 11 also includes a summary of option rankings based on

average incremental costs, to help explain why different options are being selected compared to

others.

3.1 Refinement of the strategy from draft to final plan

The draft WRMP strategy is published for consultation with customers, stakeholders and regulators.

The responses received during the consultation may result in changes to the assumptions or inputs

used to derive the supply-demand balances, as well as to the set of options that are available to

meet forecast deficits. The development of the plan as presented in the final WRMP is thus an

iterative process, in which the above decision making approach is repeated and refined in production

of a revised draft WRMP and then a final plan following consultation on the draft WRMP.

Consequently, elements of the strategy may need to be reviewed or refined as we move from the

draft plan to our final plan. The process that we followed for the production of our WRMP is

summarised below.



Figure 8 Development of the strategy from draft to final plan

3.2 Inclusion of customer preferences The company has undertaken a process of engagement with stakeholders and customers to learn

about their priorities, seek views on the development of our plans, find opportunities for collaboration

and learn from examples of best practice. This formed a key part of the process of preparing the

plan, and is described fully in Annex 1.



The ‘pre-consultation’ process also included engaging the regulators to keep them informed on the

developments of our plan, to explain our proposed methods and approaches, refine these depending

on feedback received, report results, and provide the regulators the chance to engage and influence

the development of the plan throughout the process.

The pre-consultation survey and a second survey of customers preferences following completion of

the draft WRMP were important to better understand customers’ views. In addition, consultation

representations were received from a range of stakeholders and regulatory bodies.

Both quantitative and qualitative analysis was conducted. The results of the quantitative survey

conducted during consultation on the draft WRMP are presented in the table below. This shows the

ranking of schemes by customer preference, based on the average that customers increased or

decreased the amount of an option in comparison to other option types – i.e. they could adjust the

composition of the solution from the default draft WRMP solution setting. The costs of these

adjustments were notified to them in terms of the impact on an average household bill. The column

for average bill change therefore shows the average of customers’ relative acceptance of changes

to their bills from the adjustments they were making to the option compositions, across all

respondents. This gave an inferred ranking of the option types.

Table 2 Results of quantitative survey of customer preferences

Some key aspects that this analysis highlighted were:

◼ No options were rejected

◼ Many of the selections made by customers in terms of the optimum mix of options matched

closely to our draft WRMP strategy. This suggested that customers broadly supported the

strategy in terms of the mix of options. Some framing affects were noted, where we had

already used all the available capacity of an option in the draft WRMP default setting, such

as catchment management, their contribution was reduced to allow other options to appear

◼ Tariffs, new reservoirs and water saving devices had the biggest increases in terms of the

selection of additional capacity, but these were not included in the draft plan because of the

pre-consultation responses and better alternative solutions



◼ Water saving devices were used more extensively in the preferred strategy. This is contrary

to the findings of the pre-consultation surveys which indicated that customers were not willing

to let people in to fit devices to their appliances. This later point has also been noted in pilot

studies that we have run before

◼ Options for water trading were generally reduced in terms of capacity by customers,

indicating that they did not favour water trading as much as other in-house options, which is

contrary to emerging regulation and policy in this area

The qualitative research may be summarised in general:

◼ Reducing leaks generates a desire for maximum investment in all areas and it is seen as a

key priority for Southern Water to reduce its leakage level

◼ Desalination is a proven technology, which is understood to be widely used elsewhere in the

world. More environmentally sophisticated areas (Brighton) are wary of high running costs.

New reservoirs are seen to generate really attractive environments for people and social

assets e.g. fishing, dog walking, sailing etc.

◼ Catchment management has a positive environmental / resilience benefit, so seems like a

sensible thing to do

◼ Underground storage is a popular choice

◼ Low cost of helping customers use water more wisely generates maximum investment

available in Southampton and Brighton in particular

◼ Tariffs are seen as being over complex and would be responded to negatively by bill payers.

◼ Trading water was not seen as a long term viable solution. Water saving devices are seen to

be incredibly expensive

The findings from the qualitative research sometimes differed from the quantitative research, while

the pre-draft consultation results differed in some respects from the surveys conducted during

consultation. This is an interesting finding in itself – that customer preferences are not universal. The

key areas of divergence between the quantitative and qualitative research are summarised below.

Table 3 Key areas of divergence between quantitative and qualitative research

Option type Quantitative Qualitative Our response

Tariffs Would like to see

more of these in a

future strategy

Do not like the idea

of penalty type

options

Continue to trial incentive-based

tariffs and undertake further work

in this area

Catchment

management

Reduced the overall

level of catchment

management

schemes in the plan

Think these are a

good idea

We have continued to maintain

the level of catchment schemes in

the plan as they form a cost

effective solution for customers.

They also align with regulatory

expectations.

Trading water Reduction in the

volume of water to be

relied on for future

solutions

Trading water not

seen as a long term

viable solution

Both of these opinions are counter

to government policy. We will

continue to develop a network to

promote increased resilience

across the South East region

Water saving

devices

Would like to see

more devices in the

plan

Water saving devices

are seen to be

incredibly expensive

Our ‘Target 100’ policy sets out a

broad strategy in which a range of

options can be offered to

customers



These studies and representations, and those from the previous WRMP (published in 2014), have

informed the development of the company’s stance on appropriate levels of service and, together

with feedback from stakeholders, has helped us to understand views and preferences on the supply

and demand management options that make up our options set. It has been applied to the

development and formulation of our preferred strategy by excluding options that were not likely to

meet customer or regulator expectations in the options appraisal. Where there are some differences

in the outcomes from different customer research we have set out our proposed way forward which

either involves aligning with Government ambition, regional strategies or the customer position with

a provision to gain further insight to help deliver some of these options.

3.3 Environmental cumulative impact assessment and programme appraisal

A detailed environmental assessment (covering Strategic Environmental Assessment (SEA),

Habitats Regulations Assessment (HRA) and Water Framework Directive (WFD) assessment) was

carried out on the wide range of feasible options considered for inclusion in the strategy for each

area to help inform decision making on the final strategy. In particular, the findings of the feasible

option assessments were used to evaluate the environmental and social performance of a range of

alternative strategies for maintaining a supply-demand balance in each area, with each alternative

strategy comprising a different mix of options and option types.

For each alternative strategy, the likely scale of adverse and beneficial environmental and social

effects for each option was considered, both on its own and in combination with the other options

included in that strategy. The potential effects in combination with any other relevant projects, plans

or programmes (for example, any planned major infrastructure schemes that may be constructed

and/or operated at the same time and affect the same environment and/or communities) was also

assessed. This appraisal of each alternative strategy also included consideration of the potential for

any regulatory compliance risks associated with the Habitats Regulations and Water Framework

Directive.

The environmental and social performance of each alternative strategy was used to help make

decisions on which strategies to explore further through the Real Options modelling process and to

finally determine the appropriate strategy for inclusion in the WRMP. Where appropriate,

modifications to the potential strategy were made as part of this process where environmental and

social effects were considered challenging.

As well as the adverse effects of options, we also looked at the beneficial effects of options to identify

whether any options should be prioritised in view of the environmental or social benefits they may

bring.

Once the final strategy had been determined, a final environmental assessment was carried out to

examine whether there were any cumulative effects from construction and/or operation, and whether

further mitigations measures may need to be adopted.

3.3.1 Environmental and social valuation

The company has not quantified the environmental costs and benefits of options in monetary terms

(this approach is in accordance with the EA supplementary guidance note on environmental

valuation, Nov 2016). Instead, the potential environmental impacts of options have been assessed

through the Strategic Environmental Assessment, Habitats Regulation Assessment screening, and

Water Framework Directive assessments, as outlined above. We also describe, in section 8.5 of

Annex 14, how the SEA maps to an ecosystem services approach. We have used the SEA to provide



the quantitative and qualitative assessment of our options and plan, and this has informed the

decision making of the preferred plan (as we describe in Annexes 9-11).

This approach avoids the potential for ‘double counting’ environmental costs/benefits, and the

difficulties historically experienced in quantifying the costs of environmental impacts. It is widely

acknowledged that much of the valuation evidence included within the Benefits Assessment

Guidance (BAG) is now relatively dated; however, the availability of suitable up-to-date studies is

limited. The company is aware that this is a developing area and new Natural Capital valuation and

accounting tools are under development for the water industry. These tools were not available to

inform this plan; however, it is expected that suitable valuation approaches that are currently under

development could be used for WRMP24.

3.4 Regional planning We are part of the WRSE regional planning group, a sector-wide partnership that develops a south-

east strategy for water every five years. The core membership comprises six water companies

(Affinity Water, South East Water, Southern Water, SES Water, Portsmouth Water and Thames

Water) working alongside the Environment Agency, Ofwat, the Consumer Council for Water, Natural

England, the Department for the Environment, Food and Rural Affairs (Defra), the Canal and River

Trust, the Greater London Authority, and other partners.

The aim of WRSE is to identify how best to share the water resources at a regional level. It also

looks further afield, working with neighbouring regions of the UK and their water companies to

explore the potential for inter-regional water transfers.

Our work focuses on exploring opportunities across the region for existing and new water resources

to be shared in the most efficient way to provide reliable, sustainable supplies across the region, at

best value to customers while also protecting the environment. This is because we expect the

pressure on water supplies in south east England to increase in the future due to many reasons

including climate change, population growth and the need to further protect the environment.

The water supply network within south east England is a complex pattern of different water company

areas and WRZs. This is a result of the historic development and integration of local systems over

more than a century, plus the fact that division of the region after privatisation did not necessarily

align with catchment or water resource system boundaries. Therefore, the fundamental approach

of the WRSE group is to ignore water company boundaries, to assess best ways to share available

water from the perspective of maximising regional resilience.

Many of the WRZs across the South East currently, or in the future, will experience shortfalls in water

availability. However, there are also areas that have water that can be shared. By looking at a

regional scale we can try to maximise the benefits of sharing of water resources across the area,

and in doing so, reduce the need for new water schemes or developments, and / or reduce existing

abstraction.

Our planning work helps us to understand which options might be best for the South East in the long-

term (such as identifying strategic schemes that can be optimised to provide benefit on a regional

scale), which will help the region become more resilient to drought, outage and environmental risks.

For the PR19 planning cycle, WRSE looked over a long horizon of sixty years (from 2020 to 2080)

exploring a range of different factors, including a greater range of future droughts of differing

severities, different population growth forecasts, resilience to extreme events, and reducing water

demand and leakage rates still further.



To inform the draft WRMPs, the WRSE group examined nine potential futures using an Economics

of Balancing Supply and Demand optimisation model, selecting from ~1400 options to see what

groups of options were the best choice to satisfy the deficit, and to test their resilience. This approach

is more sophisticated than an average incremental cost approach (stacking the cheaper unit cost

schemes together), but not as advanced as the Real Options approach adopted by Southern Water.

Following the close of the consultation period on the draft WRMPs, further regional modelling was

undertaken, exploring more scenarios to assess the feedback from customers. In addition, the

scenarios explored included a range of regional targets to assess the effect of meeting the

recommendations from the National Infrastructure Commission and Defra on leakage and per capita

consumption in terms of option selection.

Southern Water’s WRMP is based on a higher water efficiency scenario, which is tested against a

range of sustainability reductions set against a range of droughts up to and including a 1:500 year

design drought, allowing for drought permits and orders to be used. Therefore, there is not an exact

comparison / match with one of the WRSE scenarios.

Nevertheless, an analysis of individual company plans has shown that there is a high uptake of the

WRSE regional outputs into company preferred plans; the precise number depends up on which of

the scenarios are being compared with the company’s WRMP. Transfers of water within the region

contribute the most (mostly over 40%) to satisfying the regional water deficit.

There are two key aspects of regional planning that are particularly relevant to the preferred plan.

The first is the potential for joint schemes – for example there are two potential schemes for the joint

development of a water reuse option with South East Water. The second, and perhaps more

common example relates to water trading – i.e. bulk supplies with neighbouring water companies.

Water trading can provide greater resilience in the supply system which customers support.

However, there can also be a number of limitations to water trading, such as:

◼ The timing by which transfers are available to provide supplies, compared to when we

actually face a deficit that needs to be resolved

◼ The security of the supply of traded water, and particularly whether the source used to provide

the supply could be at risk of a future sustainability reduction

◼ The extent to which the supplying company can provide a guarantee that the water supply

will be available during the drought return periods that we are planning to

◼ The cost of the bulk supply – bulk supply options must be economic in comparison to our

own resource development options, as it would not benefit our customers if trading options

were significantly more expensive than our own options. This may occur where, for example,

the supplying company charges for the development of a new water source and we then also

face the cost of the pipeline from that source to our own supply area whilst also paying

relatively high costs per unit of water supplied

◼ It is also important to understand that bulk supply agreements cannot be completely reliable

in all drought events, as the donor company has a duty to maintain supplies to its own

customers

Nevertheless, the company continues to discuss and explore water trading options with neighbouring

water companies. We have led the development of more robust water resources planning by

introducing stochastic modelling into the sector to provide insight into potential future droughts, and

the use of real options methods to provide adaptive and scalable solutions. In the future the remit of

the WRSE could be extended such that they would derive a regional plan that would then be provided

to the water companies to incorporate into their business plan.



We are committed to support the development of a regional water transfer grid that will support water

trades between companies and increase the level of resilience to drought and other incidents for

customers in the South East. In Annexes 9, 10 and 11 we have provided further information on how

our preferred strategies in each area support the development of this regional grid.

3.5 Development of strategies for each area From the initial formulation of a constrained least cost strategy, through testing the plan, a portfolio

of options (the strategy) was identified, as described previously. This differs from the least cost

solution, as it takes account of other criteria and policy decisions to ensure that the plan represents

the optimum balance of monetary and non-monetary issues, risks and uncertainties, and customer

preferences.

The analysis also helped to identify the key investigations that will be required over the next 5 to

10 years to ensure that options can be introduced in a timely manner when required.

The strategies for each area, including key investigations and any associated risks and uncertainties,

are presented in separate Annexes:

◼ Annex 9: strategies for the Western area

◼ Annex 10: strategy for the Central area

◼ Annex 11: strategy for the Eastern area

This plan presents an update of the strategies for each of the three supply areas, following

consultation on the draft WRMP (which ended on 28 May 2018).

3.5.1 Options and strategy risks

Each strategy comprises a portfolio of options – a combination of resource development, demand

management, bulk supplies and inter-zonal transfers. However, some options may present greater

risks than others in terms of how easily they may be implemented due to, for example, planning or

environmental risks, or technological complexity.

Therefore, Southern Water has sought to understand these risks and identify alternative options that

may be required should the risks of a given option materialise. This was through a process of

scenario and sensitivity testing, as discussed in section 3.6 below.

Annexes 9-11 present timelines for the key strategic options to demonstrate issues which could

impact on the deliverability of the schemes. However, there will always be some deliverability risk

where new large-scale water resource options may need to be developed. As a result, the timelines

also show the alternative schemes that would get triggered in the event of a strategic scheme being

undeliverable, and the impact this could have on the timing of the plan.

These timelines also highlight key decision points or triggers which might result in a change to a

different ‘branch’, or a change of strategic scheme (i.e. remove the need for a scheme, trigger the

need for a new or larger scheme, or a move to an alternative scheme).

Uncertainty in the supply and demand forecasts will increase through time, and so it is logical to

divide the water resource strategy into key time periods over the planning period:

◼ The first identifies the next strategic schemes required for which funding for implementation

must be sought through the forthcoming Business Plan (covering the period 2020-25)



◼ The second identifies those schemes for which implementation is not required in the next 5

years but which will require further investigation (because they may need to be implemented

in the next 10-15 years) to ensure that they are feasible before the next WRMP is produced

in 2022-23, and to ensure that any required planning permissions can be obtained and any

environmental issues can be addressed and mitigated

◼ The third considers options that may well be required in the longer term. The purpose is to

understand the strategic nature of schemes which may be required in the longer term, but

which are subject to greater uncertainty and will need to be confirmed or revised in

subsequent WRMPs

3.5.2 Resilience

Some options may provide improvements to the supply system’s resilience to different drought

events or other unplanned outages; i.e. they may provide the company with greater flexibility to

respond to a range of unforeseen events.

There are four components of resilience: resistance, reliability, redundancy, and recovery. In very

general terms, the WRMP provides more of a resistance component of resilience, while the recovery

aspect of resilience is important for the environment after a drought event. The aspects of reliability

and redundancy are also addressed in WRMPs – for example, through activities to minimise outages

and options to refurbish/rehabilitate source works, and through greater connectivity between

resource zones and through trading with neighbouring water companies.

Our plan focuses on a wide range of resilience measures:

◼ We have developed an approach that solves multiple drought events and inter-annual

variability simultaneously. This includes assessment of extreme drought conditions, which

aligns with the recent NIC3 report, to ensure we have a plan that is resilient to drought events

to minimise the potential for ‘level 4’ type restrictions such as standpipes and rota cuts. These

can have significant impacts on society and the economy

◼ Our demand management activity in the last AMP, and as proposed in this current plan will

also contribute to our resilience to drought events. For example, during the summer 2018

heatwave, our peak week distribution was about 100Ml/d lower than the last really hot, dry

summer spell in the South East in 2003. This represents demand reduction of around 15%

against the 2003 peak week demands. We believe that this is driven largely by our high

penetration of domestic metering, and associated water efficiency activity, in addition to

ongoing leakage reduction activity. Our plan seeks to reduce demand further in future, which

should continue to provide resilience to hot, dry weather events

◼ We have a strong focus on water trading in our plan in line with regional planning outcomes,

in addition to improving the connectivity between our WRZs. In general, greater connectivity

will provide greater resilience, and therefore reduce risks from outage and events such as

extreme droughts, heatwaves, freeze-thaw, pollution or even terrorism.

◼ The above will contribute to increased resilience across the whole south east region,

particularly where bulk supplies are designed to be bi-directional

◼ Our outage assessment, described in detail in Annex 3, has included identification of water

quality risks from Drinking Water Safety Plans

◼ Our plan aims to reduce outage by the end of AMP7, and therefore to adopt measures,

through the business planning process, to increase system resilience to outages

◼ We also have a wide range of catchment management schemes focused on increasing the

resilience of our supplies (in combination with improvements to the treatment works) from

nitrate and pesticide risks

3 National Infrastructure Commission (April 2018), Preparing for a drier future: England’s water infrastructure needed



◼ In addition to the water quality-focused catchment management measures in our plan, we

have also included in-stream improvements and mitigation measures associated with the

Lower Test and Itchen rivers. These measures together are aimed at improving

environmental resilience

◼ We have tested our plan against winter peak demands, which are generally driven by freeze-

thaw events which drive short term increases in leakage. A recent example of a significant

freeze-thaw incident was experienced in early 2018. We have provided analysis of how our

sources and the schemes selected in our preferred plan perform against freeze-thaw events

for each area (in Annexes 9-11).

◼ The company has completed an assessment of all its water supply assets to identify those

at risk of flooding events. Three sources were found to be at risk of a 1:200 year flood event,

and as a result works are being progressed currently to address this at these three sources.

Flooding risk is also included as part of the SEA assessment criteria in the assessment of

our preferred options, and these options will be designed according to national Ofwat

planning guidance relating to flood risk.

The company has documented those options which may provide additional resilience benefits in

each of the Annexes 9-11 for each supply area.

3.6 Sensitivity testing of the strategy Having developed the strategy or plan for the WRMP through testing a constrained least cost

strategy against SEA criteria, WRSE regional plans, and customer preferences, as described

previously, the next step is to undertake sensitivity testing of the strategy. The process of stress

testing and sensitivity analysis helps the company to identify and understand the assumptions and

factors that have the greatest influence on the plan, the key decision points, and range of potential

alternative options, and thus ensure that the plan is robust under a wide range of uncertainties.

A Real Options modelling approach already allows for uncertainty around how different futures may

evolve and thus trigger different options selection. Our approach therefore already provides some

evaluation of alternatives in the strategy and therefore reduces the requirement of sensitivity analysis

to some degree (UKWIR, 2016).

Nevertheless, sensitivity testing was performed on the plan. The purpose of sensitivity testing is

twofold:

◼ To ensure the plan is as robust as possible in the face of uncertainties. This provides

confidence in the portfolio of schemes selected;

◼ To understand the range of potential alternative options if preferred options cannot be

implemented for whatever reason. This may require feasibility studies, investigations or

planning activity to be carried out in parallel to the main portfolio of options in the strategy.

We developed a range of sensitivity testing model runs to compare against the strategy. The

sensitivity tests could be specific to a given area, or applicable to all three areas. The sensitivity

testing that was applied to each area is described in Annexes 9 to 11. The range of sensitivity tests

applied was also refined as we progressed from the draft plan through to the final plan.

In general, we examined:

◼ Uncertainties around particular options or bulk supplies, or around policy assumptions

◼ Assumptions around Drought Permits/Orders to relax abstraction licence conditions, and

their availability in severe droughts

◼ Cost uncertainties



◼ ‘Accepting’ deficits for the first two AMP periods to confirm that the options selected in the

strategy are indeed optimal, and their selection is not driven purely by them being available

for delivery before other options

◼ The impact of sustainability reductions

◼ Alternative outage allowances

◼ Comparison to a conventional modelling approach i.e. EBSD

◼ Comparison to the strategy developed for the previous plan, WRMP14 – which was based

on a conventional EBSD modelling approach, but which would also have had different

assumptions regarding the supply and demand components (and therefore have been a

slightly different problem to solve)

◼ The impact of different branch weightings

The sensitivity testing allows the company to understand whether there are key alternative strategic

schemes which may be needed, should it not be possible to implement the schemes in the preferred

plan. This is particularly important for those schemes in the strategy that are required in AMP7 or

AMP8.

Where there may be some uncertainty around the delivery of these schemes, the company may

need to conduct feasibility investigations of alternative schemes (and potentially environmental

surveys and planning activities) in parallel to developing the portfolio of schemes selected in the

strategy.



4. References

◼ Environment Agency and Natural Resources Wales, April 2017, Water Resources Planning

Guidance and supporting documentation: Interim update

◼ HM Government, Treasury Green Book

◼ National Infrastructure Commission (NIC) (April 2018), Preparing for a drier future: England’s

water infrastructure needs.

◼ Southern Water, 2014, Water Resources Management Plan 2015-40 Technical Report,

Southern Water Services, Worthing.

◼ UKWIR, 2016, WRMP19 Methods – Decision making process: guidance, UKWIR Report Ref

16/WR/02/10

◼ UKWIR, 2016(b), WRMP19 Methods – Risk based planning methods, UKWIR Report Ref

16/WR/02/11

Water Resources Management Plan 2019 Annex 8: WRMP Strategy

Appendix A: WRMP Tables December 2019 Version 1


Annex 8: WRMP Strategy Appendix A: WRMP Tables

RESTRICTED INFORMATION, AVAILABLE UPON REQUEST

Date post:	17-Mar-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

Water Resources Management Plan 2019 Annex 8: WRMP strategy · 5 Water Resources Management Plan...

Documents